Mehdi F.

Led AI R&D of an internal product based on LLMs like GPT-4 and Claude for NLP and AI applications.

Worked on LLM and application evaluation / assessment.

Integrated frameworks like LangChain to build advanced RAG applications and improve model performance.

Provided technical leadership, MLOps best practices, and DevOps workflows.

Led cross-functional teams in the design and implementation of microservices architecture on cloud platforms, leveraging Google Cloud services. Spearheaded the development of a backend infrastructure using FastAPI, ensuring high performance and API endpoints for data exchange. Orchestrated integration of machine learning resources into the application stack, enabling predictive analytics and enhancing user experiences. Implemented CI/CD pipelines ensuring rapid development cycles. Led the adoption of containerization and orchestration technologies, including Docker and Kubernetes, optimizing resource allocation and scalability.

Developed state-of-the-art models to monitor vital signs, driving innovation and improved patient care. Spearheaded the development and implementation of multi-modal deep learning models using TensorFlow, PyTorch, and TensorRT, contributing to cutting-edge healthcare technology. Co-authored a patent proposal in vital-sign monitoring. Innovated and implemented machine learning pipeline tools, including DVC (Data Version Control) and CML (Continuous Machine Learning), streamlining data processing and model deployment for enhanced efficiency and accuracy. Leveraged ONNX for model interoperability, enabling integration across various platforms and frameworks. Leveraged TensorRT for execution optimization. Utilized Google Cloud to accelerate model development and deployment, optimizing scalability and performance. Collaborated closely with cross-functional teams to integrate the solutions into existing workflows, resulting in significant performance improvements in model training and deployment.